Add http proxy settings to the config file.

Bug #235497 reported by Herb McNew
14
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Launchpad itself
Fix Released
Low
Colin Watson

Bug Description

In order to connect to outside websites (like Google) we need to use an http proxy from the datacenter. In order to assure things like search and bug watches work we need to be sure to set the http_proxy environment variable. It would be better if that information was stored in the config system for launchpad itself so it wouldn't have to pick that up from the environment.

Fixing this bug would also fix bug 235275

Changed in launchpad:
importance: Undecided → Medium
status: New → Confirmed
Revision history for this message
Diogo Matsubara (matsubara) wrote :

Rationale behind this bug:

<flacoste> matsubara: actually, we don't need a patch because the LOSA updated the staging and production update scripts to set that variable
<flacoste> matsubara: the bug is to have a generic fix for all cases where this might happens
<flacoste> the bugwatch script also needs that variable for instance
<flacoste> we could probably have an 'environment' section where all keys are set in the env
<flacoste> at startup

Without the env variable set we end up seeing OOPS like: OOPS-879S23
Exception type URLError
Exception value <urlopen error (111, 'Connection refused')>

Revision history for this message
Guilherme Salgado (salgado) wrote : Re: [Bug 235497] [NEW] Add http proxy settings to the config file.

We already have config.launchpad.http_proxy. What else would we need?

 subscribe

Revision history for this message
Francis J. Lacoste (flacoste) wrote :

We should probably configure the environment variable automatically based on that when canonical.config is loaded.

description: updated
Curtis Hovey (sinzui)
visibility: private → public
tags: added: canonical-losa-lp
Curtis Hovey (sinzui)
Changed in launchpad:
importance: Medium → Low
Revision history for this message
Colin Watson (cjwatson) wrote :

A while back (2018 or so) I went through and converted all our external HTTP calls to lp.services.timeout.urlfetch so that we had a central place to ensure that they handled proxies correctly, and then arranged to set the proxy only in lp:lp-production-configs and not in cron job environments. That fixed this bug.

Changed in launchpad:
assignee: nobody → Colin Watson (cjwatson)
status: Triaged → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.